The successful identification of numerous T cell epitopes derived from tumor-associated antigens has opened the doors to the development of subunit vaccines for the treatment and prevention of cancer. Vaccines prepared from synthetic peptides representing these T cell epitopes constitute an attractive approach for the induction of anti-tumor immune responses because they are easily manufactured, simple to characterize, relatively stable and most importantly, extremely cost effective as compared to other types of vaccines. However, because of MHC restriction, a major drawback of peptide epitope based vaccines is that they offer limited population coverage. Moreover, since one single tumor antigen will not target all patients with one type of malignancy, T cell epitopes from several antigens will be needed to produce vaccines for broad use in each disease setting. During the 1st funding period, we have focused on the identification of peptide epitopes for cytotoxic T lymphocytes (CTL) and helper T lymphocytes (HTL) from CEA, MAGE3 and HER2/neu, which are commonly found on epithelial cancers. These epitopes could be used for developing peptide-based therapeutic vaccines for breast cancer. However, the use of peptide mixtures, or cocktails in clinical studies is a complex endeavor that is not only costly, but also frowned upon by regulatory agencies. Each peptide has to be produced separately under good manufacturing practices (GMP) and has to undergo individual quality control testing, which is very costly. Responding to this problem, we have designed peptide-based vaccines, which we named "Trojan Antigens" (TA) that allow the use of multiple T cell epitopes in a single construct. Using murine T cell epitopes we have shown that the TA can generate MHC:peptide complexes in a proteosome/TAP independent fashion and are capable of inducing strong immune responses in vaccinated mice. For the present grant, we would like to apply the technology of TA to human CTL and HTL epitopes and evaluate their immunogenicity in HLA transgenic mice. These results will serve as preclinical data for the initiation of clinical trials in cancer patients. The following specific aims are proposed: 1) To evaluate the immunogenicity of TA containing multiple CTL epitopes from CEA, MAGE3 and HER2/neu in HLA-A2 transgenic mice. 2) To evaluate the immunogenicity of TA containing CTL and HTL epitopes from CEA, MAGE3 and HER2/neu in HLA-A2/-DR4double transgenic mice. 3) To study anti-tumor immunity of TA containing multiple CTL and HTL epitopes in HLA-A2/-DR4 mice expressing the ratHER21neu oncogene. We believe that by completing the above specific aims, we will generate sufficient preclinical data to take this approach into the clinic.